Co-occurring pain, fatigue, and sleep disturbance is a common symptom cluster among patients with advanced cancer, diminishing functional status and quality of life. Cognitive-behavioral (CB) strategies (e.g., relaxation, distraction, and imagery) are effective treatments for each of these symptoms, and provide a logical choice in managing all three co-occurring symptoms. CB strategies are thought to work psychologically by reducing anxiety and stress, improving expectations for symptom outcomes, and enhancing perceptions of personal control. Recent research also suggests a physiologic pathway, in which CB strategies attenuate dysregulation in stress hormones and reduce proinflammatory cytokines that contribute to symptom exacerbation. CB strategies are not equally effective for all patients, with differences based on individual preferences and skills and concurrent symptoms. Thus, training in a single strategy is not sufficient. Our previous work demonstrates feasibility and initial efficacy of a patient-controlled cognitive-behavioral (PC-CB) intervention in which patients receive information about their symptoms and rationale for CB strategies, then self-administer their selection of various relaxation, distraction, and imagery strategies via MP3 player, using the CB strategies at whatever time and place they are needed. This brief intervention was designed to be mindful of the high symptom burden, unpredictable treatment course, and often short life expectancy of persons with advanced cancer. The purpose of this study is to test efficacy of the PC-CB intervention in a RCT and to test moderators and mediators of intervention effects. The specific aims are (1) to test efficacy of the PC-CB intervention on severity, distress, and interference with daily life from the pain, fatigue, sleep disturbance symptom cluster, (2) to test proposed moderators (imaging ability, concurrent symptoms) of PC-CB intervention effects, (3) to test psychological mediators (anxiety, stress, outcome expectancy, perceived control) of PC-CB intervention effects, and (4) to explore two types of physiological mediators (neuroendocrine stress hormones and indices of inflammation) of PC-CB intervention effects. A sample of 160 adults receiving chemotherapy for advanced breast, lung, colorectal, prostate or GYN cancers and experiencing pain, fatigue, and sleep disturbance will be recruited. Patients randomized to the PC-CB intervention will receive a brief training session and instructions to use the recorded CB strategies during 9-weeks of cancer treatment. Patients randomized to attention control will receive instructions to use cancer educational recordings over the same period. Measures of symptom severity and distress, symptom interference with daily life, and proposed psychological and physiologic mediating variables will be completed at baseline and the 3-week primary endpoint. Symptom outcome measures will be repeated at 6- and 9-week secondary endpoints to evaluate longer term treatment effects. If the PC-CB intervention is efficacious, it could be readily translated into practice with little burden, and would move the field toward a more pragmatic approach to treating multiple, clustered symptoms.